Catalytic Oxidation of Ethylbenzene in the Liquid Phase - Industrial

1 May 2002 - Catalytic Oxidation of Ethylbenzene to Acetophenone Using Alumina-Supported Dichromate: Process Optimisation and Development of a ...
2 downloads 0 Views 300KB Size
1N D U S T R I A L A N D

1286

E N G I N E E R I N G C H E 21. I S T R Y

Makowwiecki, J. Russ. P h y s . Chem. Soc., 40,752 (1908). Mason a n d Wheeler, J . C h e h . SOC.,121, 2079 (1922). Masson a n d Hamilton, IND.E m . CHEM.,19, 1336 (1927). Ibid., 20, 813 (1928). Moore, J. SOC.C h e m . Ind.,36, 109 (1917). Paterno and Spallino, Atti mad. Lincei, [ 5 ] 16, I 87 (1906). ESG.CHEM.,21, 695 (1929). Reid and Hoffman, IND.

(15) (16) (17) (18) (19) (20) (21)

Vol. 25, No. 11

(22) Taffanel and Le Floch, C o m p t . rend., 157, 469 (1913). (23) T a n a k a and S a g a i , Proc. I m p . Acad. (Tokyo), 2, 219 (1926). (24) T a n a k a and Sagai, J. SOC.C h e m . Z n d . Jagon, 29, 266 (1926). (25) Thompson, IND. ENG.CHEM.,21, 134 (1929). (26) Wurtz, Ann. chim., [3] 69, 323 (1863). R E C E I V EM D a y 23, 1933. Published by permission of the Director, U. S Bureau of Mines. (Not subject to copyright.)

Catalytic Oxidation of Ethylbenzene in the Liquid Phase ~

AND J. J. STUBBS, Bureau of Chemistry and Soils, Washington, D. C. C. E. SENSEMAN

N A PREVIOUS article (4)

I

The catalytic action of manganese dioxide in ber, which was then placed in the authors presented data a glycerol b a t h p r e v i o u s l y the liquid-phase oxidation of ethylbenzene has on the liquid-phase oxidaheated to approximately the been studied, with particular emphasis on the tion of p-cymene in which a desired temperature, and the formation of acetophenone. Phenylmethylcarfinely divided manganese dioxflow of oxygen w a s s t a r t e d . binol is found as a reaction product in almost ide prepared by reducing potasAfter the oxidizing period, the sium permanganate with formalweight of the material in the constant quantities after oxidation periods of dehyde was used as a catalyst. reaction chamber was obtained, 5.75 hours or more. An explanation is adIt was shown that this oxide as well as that of the products vanced f o r this constancy of yield. Reaction had a decidedly accelerating which had distilled off during products identijied were: wafer, carbon dioxide, effect upon the reaction. the course of the run and had formaldehyde, phenylmefhylcarbinol, acetopheAfter that investigation it was been condensed and collected decided t o make a similar in a flask immersed in an ice none, and benzoic acid. study of the catalytic value of bath. this same oxide in the oxidation ANALYSIS O F P R O D U C T S . of ethylbenzene, looking particularly toward the production Oxidation products detected qualitatively were: of acetophenone, an organic chemical of increasing imporIDENTIFICATION METHOD tance. I n studying the mechanism of the oxidation reactions Phenylmethylcarbinol M. p. of the acid phthalic ester, 108-108.5° C. neutralization equivalent of thla ester, (cor.) ; of a number of hydrocarbons, Stephens (5) found that aceto267 M. p. of the phenylhydrazone 102-103' C. (cor.) phenone is produced in considerable quantity when oxygen Acetophenone M. p., 121.20 C. (cor.); nedtralization equivais passed through ethylbenzene over a long period of time and Benzoic acid lent. 123 Formaldehyde R&g&oitest no catalyst is used. King, Swann, and Keyes (3) found that Carbon dioxide Barium hydroxide teat Cupric sulfate teat manganese acetate catalyzed this reaction to the extent that Water 20.6 per cent of the ketone, based upon the initial material, The phenylmethylcarbinol, acetophenone, and benzoic was formed during a period of 24 hours, the reaction being carried out a t a temperature of 102' to 104' C. Keither of acid were determined quantitatively as follows: these investigations showed the formation of the alcohol, The solution of these three in ethylbenzene as taken from the phenylmethylcarbinol, or the further oxidation of aceto- reaction chamber was thoroughly extracted with aqueous sodium phenone t o benzoic acid. Binapfl and Krey (1) claim that, bicarbonate solution which reacted with any uncombined acid. The aqueous solution was then treated with dilute sulfuric acid after passing oxygen through ethylbenzene for 7 hours a t a and the liberated organic acid extracted with ether. After washtemperature of 130" to 140' C., using as an oxygen carrier ing the ether solution free of all sulfuric acid, it was made up to a manganous hydroxide precipitated on marble powder, 59 definite volume, an aliquot was removed, the ether was evapoper cent of the ethylbenzene is recovered and a residue rated, alcohol and water were added, and the acid was titrated with 0.1 N sodium hydroxide solution, phenolphthalein being (quantity not stated) is obtained consisting of 80 per cent used as an indicator. From this titration result, the total quanacetophenone and 20 per cent phenylmethylcarbinol. tity of acid was calculated. From the material insoluble in the bicarbonate solution, ethylEXPERIMENTAL PROCEDURE benzene was removed by distilling at atmospheric pressure from a long-neck distilling flask, after which the residue was distilled A technical grade of ethylbenzene was distilled, and the under reduced pressure to avoid possible decomposition of the fraction boiling a t 135' to 137" C. was used. The catalyst, alcohol. The alcohol fraction contained also the acetophenone. manganese dioxide, was prepared according t o the method The weights of both fractions were obtained. To a one-gram described in a former article (4). Except for the changes sample of the mixture of ketone and alcohol, acetic acid was added. The acetophenone was precipitated as the phenylhydranoted, the apparatus was also the same as described therein. zone, which was filtered off, washed with 50 per cent alcohol, For greater convenience in operation, a n accurately Cali- dried, and weighed. From this weight the total acetophenone brated flowmeter was substituted for the mechanical dry was calculated. The alcohol in the mixture was determined by meter, and a sintered disk made of powdered Pyrex glass was the acetylation method of Gildemeister and Hoffmann (9). sealed into the bottom of the reaction chamber as a substitute DISCUSSION OF RESULTS for the alundum disk. This glass disk gave better dispersion Table I has been compiled primarily to show the effect of of the oxygen gas, resulting in better contact of gas, liquid, the temperature variable. B y the inclusion of the first and and catalyst, and better agitation. I n conducting the experiments, 50 grams of ethylbenzene last runs given, comparison is also made between results oband 0.5 gram of catalyst were poured into the reaction cham- tained when using no oxide, a finely divided commercial man-

,/

-'

November, 1933

IKDUSTRIAL AND ENGINEERING

CHEMISTRY

1287

TABLEI. TEMPERATURE EFFECT (50 grams ethylbenzene; 0.5 gram catalyst; length of run, 5.75 hours; 125 cc. of oxygen per minute) ETHYLBENZENE YIELDSOF OXIDATION PRODKXTS~.-ReUnrePhenylmethylRUN TE'hfP CATALYST covered covered Acetophenone Benzoic acid carbinol c. Grams Grams Grams % Grams % Grams % 1 115 None 46.0 4.0 0.6 High-boiling Trace .. Not determined fraction 2 110 MnOz 38.0 12.0 3.8 31.7 Trace .. 1.4 11.7 3 115 MnOz 32.3 17.7 9.7 54.8 Trace 2.9 16.3 4 120 MnOz 28.1 21.9 5.7 26.0 8.0 3615 2.8 12.8 5 130, 115b MnOz 25.0 25.0 15.8 63.2 1.24 5.0 3.1 12.4 6 130, 115b Commercial MnOz 45.0 5.0 0.9 18.0 Trace .. Not determined 0, All percentage yields are weight yields based upon ethylbenzene unrecovered; yields of oxidation products are calculated from analytical data. b A temperature of 130° C. was maintained for approximately 45 minutes, then lowered to 115' C.

ganese dioxide, and the specially prepared dioxide. That the latter has a decidedly accelerating effect is conclusive. Table I further shows that the rate of attack upon the ethylbenzene molecule, as well as the relative amounts of acetophenone and benzoic acid formed, is greatly influenced by small temperature variations. I n both Tables I and I1 the amount of ethylbenzene unrecovered in each run was arrived a t by subtracting from the weight of the charge (50 grams) the weight recovered by distillation after acid extraction, plus the weight of the hydrocarbon distilling over during the run and collected in the flask in the ice bath. The inherent losses of this procedure account for the relatively high amounts of ethylbenzene designated as unrecovered, as compared with the total oxidation products recovered. The percentage yield of each oxidation product was calculated as a weight yield based upon the weight of ethylbenzene unrecovered. TABLE11. TIMEEFFECT (50 grams ethylbenzene; 0.5 gram catalyst; 125 cc. of oxygen per minute. temperature, 130' C. for approximately 45 minutes and 115' for remainde; of the time) --YIELD0 OB OXIDATION PRODUCTSETHYLBENZENE PhenylReUnreAcetoBenzoic methylRUN TIME covered covered phenone acid carbinol Hours Grams Grams Grams % Grams 70 Grams % ' 1 4.5 42.0 8.0 2.2 27.5 Trace 1.2 15.0 2 5.75 25.0 25.0 15.8 63.2 1.24 510 3.0 12.4 3 8 25.5 24.5 11.8 48.1 4.10 16.7 3.0 12.2 4 loa 23.8 26.2 14.4 55.0 6.44 24.5 2.5 9.5 a The temperature was 115' for the entire period.

While making various runs a t a given temperature for the entire period, it became evident that an appreciable induction period was necessary, after which the reaction proceeded much more rapidly, this being indicated by less heat requirements and greatly increased production of carbon dioxide. I n order to lessen this period, runs were then made with an initial temperature of 130" C. until such time as the reaction became more vigorous, usually 45 minutes or more. The temperature was then lowered as indicated in Tables I and I1 and continued thus for the remainder of the period in an effort to keep down the production of acid which would have resulted a t the higher temperature. Comparison of runs 3 and 5, Table I, shows an increased production of acetophenone as a result of this procedure. A 10-hour run made a t 130" C. gave a yield of 16.6 grams of acid and 1.2 grams of acetophenone, 55.3 and 3.9 per cent, respectively, based upon the ethylbenzene unrecovered. The results obtained by increasing the oxidation periods are

AGRICULTURAL ISSECTICIDES IN MEXICO. The use of agricultural insecticides in Mexico is not extensive, and the development of the market lies in further educating the Mexican agriculturist in the importance of attempting to control insects and diseases affecting plant life. Considerable educational material is being disseminated by the Mexican Department of Agriculture with a view t o increasing the use of insecticides and fungicides, this department being thoroughly cognizant of the importance of spraying and dusting

shown in Table 11. While it is obvio6s in comparing runs 2 and 3 that oxidation progressed a t different rates, it is also evident that acetophenone production is not increased by extending the oxidation period beyond 5.75 hours. I n all runs listed in the two tables, 125 cc. of oxygen per minute were used. Other experiments showed that considerable variations in this volume affected the yields very little; there was always an excess of oxygen over that absorbed which aided in the agitation. I n both tables, when experiments were run a t 115' C. for 5.75 hours or longer, the quantities of phenylmethylcarbinol formed were quite uniform. Stephens (5) in his work on oxidation without catalysts found that when 21 grams of ethylbenzene were subjected to oxidation for 11 days a t 122" to 124" C., 6.3 grams of acetophenone were produced but no alcohol was formed. When 25 grams of phenylmethylcarbinol were similarly oxidized, only 0.85 gram of acetophenone was formed. He concluded from this, and from his results from other oxidations, that any alcohol formed is not an intermediate product but rather a by-product with formation due to some "exceptional conditions." If we assume that the catalytic oxidation is one of alternate oxidation and reduction of the metallic oxide, giving rise to momentary atomic oxygen, a very different condition exists from that which he had, and one which might produce the alcohol as a by-product. If the alcohol, being in low concentration in the reaction mixture and having, as Stephens found, a slow oxidation rate when not diluted, is very slowly oxidized to the ketone, it can be readily understood that the concentration of the alcohol might reach a point where its oxidation to the ketone would approximate its rate of formation from ethylbenzene. The concentration of the alcohol would then change but little through the remainder of the oxidation period. An explanation is thus given for the uniformity of alcohol yields, which tends toward a confirmation of Stephens' theory.

LITERATURE CITED (I) Binapfl and Krey, U. S. Patent 1,813,606 (July 7,1931). (2) Gildemeister and Hoffmann, "Volatile Oils," tr. by Edward Kremers, 2nd ed., Vol. 1,p. 575, Wiley, 1920. (3) King, Swann, and Keyes, IND. ENQ.CHEM.,21, 1227 (1929). (4) Senseman and Stubbs, Ibid., 24, 1184 (1932). (5) Stephens, J. Am. Chem. SOC.,48, 1824, 2920 (1926); 50, 2523 (1928). RBCBIVED June 20, 1933. Contribution 227 from the Color and Farm Waste Division, Bureau of Chemistry and Soils.

for the control of insects and plant diseases. In addition to general educational activities, the Mexican Department of Agriculture has also established a large fumigation plant at Guadalajara, and is understood to be contemplating the establishment of others in different parts of the country. Perhaps the principal factors t o be overcome in the development of the market are the low purchasing power of the great majority of Mexican farmers and the conservative methods employed in the cultivation of their fields.